The natural environment is cluttered with stimuli and the brain has limited processing capacity. Attentional mechanisms are therefore needed to guide the selection of behaviorally relevant information. The present application is a competing renewal submission for our project Neural basis of visual attention (R01- MH64043-10). Work during the previous funding period used fMRI to characterize attention-related functions at multiple stages of the human attention network, including the thalamus and the fronto-parietal network. The present application extends this work, using electrophysiological recordings in humans and monkeys to investigate temporal dynamics and functional interactions across the attention network (both cortex and thalamus). Specifically, we aim to characterize the neural basis of object-based selective attention. Objects are typically the units of selection. When attention is spatially allocated to part of an object, an `attentional spotlight' (and the facilitaed processing associated with it) automatically expands to match the extent of the object's boundaries. That is, attention spreads to include task-irrelevant locations within the object. Classic attention theories assume that a unitary and indivisible spatial mechanism mediates such object-based selection. However, recent evidence challenges this characterization. First, the attentional spotlight, rather than being sustained, flashes rhythmically, sampling the visual environment at frequencies in the theta band (4-8Hz), with alternating temporal windows of relatively enhanced and diminished processing. Second, our behavioral studies from the previous grant cycle support the existence of two spatial mechanisms, concurrently sampling the visual scene: (i) a fixed spotlight that rhythmically samples the most relevant location, and (ii) a moving spotlight that rhythmically monitors less relevant locations. The present project wil challenge classic models of attention by investigating the neural basis of rhythmic selective processing within the framework of object- based attention. By recording from multiple nodes of the attention network in humans and macaques (using the same attention task for both species), we will probe the central hypothesis that attentional selection is a highly dynamic process that operates concurrently at multiple locations. In humans, we will obtain intracranial recordings in epilepsy patients (i.e., electrocorticography [ECoG]). In macaques, we will simultaneously record from interconnected regions of FEF, LIP, area V4, and the pulvinar. Specifically, we will (i) investigate a dissociation of function between FEF and posterior PPC, (ii test the functional role of the pulvinar in coordinating temporal dynamics across cortical network nodes and (iii) relate neural signals, including oscillatory activity, to rhythmic behavior. Impairments of attentional selection have devastating consequences on human health (e.g., after stroke, and in diseases, such as schizophrenia). The significance of the proposed research is that it will probe potentially paradigm-shifting hypotheses using an innovative approach that combines cutting edge neuroimaging techniques with intracranial electrophysiology in two primate species.